Each year an estimated 1.3 million persons ?24 years of age receive 7 million antipsychotic prescriptions in the U.S. Although the primary indications for antipsychotics are schizophrenia and related psychoses, with no other treatment alternatives, an estimated 90% of antipsychotic prescriptions for children and youth are for other, less serious conditions, including attention-deficit/hyperactivity disorder (ADHD), disruptive or aggressive behaviors, affective disorders including bipolar disorder, and anxiety. However, other recommended therapeutic interventions for children and youth with these disorders are thought to have fewer adverse effects. Antipsychotics, which increase the risk of cardiovascular and all-cause mortality in adults, have serious adverse cardiovascular, metabolic, respiratory, and neurologic effects in children and adolescents that plausibly increase the risk of death in this population. We recently found that antipsychotic users of doses>50mg chlorpromazine equivalents (median starting dose) had a greater than 3-fold increased risk of unexpected death, leading to a 64% increase in total mortality (HR = 1.64 [1.03-2.63]). In contrast, the adjusted risk of deaths from injuries or suicides did not increase nor was there increased risk of death from any cause for lower doses of antipsychotics. Our data indicate antipsychotics increase risk of unexpected deaths, particularly cardiovascular deaths. The increased risk is clinically meaningful: the incidence of unexpected death in higher-dose antipsychotic users equaled that of injuries and suicides, which account for two-thirds of deaths in children and adolescents. Thus, death should be considered as a potential harm when prescribing antipsychotics for children and youth. However, to guide clinical practice, data are needed that define antipsychotic-related mortality: 1) according to antipsychotic indication; and 2) according to important factors that practitioners can control: a) individual drug, b) dose, and c) concurrent central nervous system (CNS) depressants. We will address these questions using the national Medicaid Analytical Extract (MAX) database, which includes more than 15 years of longitudinal data that can be linked to death certificates for the estimated 39% of children in the U.S. who are Medicaid enrollees. There are two specific aims: Aim 1: Test the hypothesis that the risk of unexpected deaths and total mortality in children and youth who are antipsychotic new users with a) ADHD or disorders of behavior/conduct, b) unipolar depressive or anxiety disorders, or c) bipolar disorders is greater than that for comparable patients starting alternative medications. Aim 2: Define how risk of unexpected deaths and total mortality in children and youth who are antipsychotic new users varies with a) individual drug, b) dose, and c) concurrent CNS depressants.